494 WHITMAN CROSS 



exerted to bring molten magmas to the sites of the rocks we 

 study, but the fact of the molten condition which gave its stamp 

 of common characters to the products of consolidation. 



Arbitrary steps are necessary in the classification of such 

 objects as rocks, exhibiting gradations in all directions. But 

 that fact does not justify such artificial systems as many 'of those 

 which have been reviewed. Among the most distinctly artificial 

 systems are those of Cordier, Senft, and von Lasaulx ; but scarcely 

 less so, as regards igneous rocks, are those which, while using 

 chemical or mineral composition as the basis of arrangement, 

 use only a portion of the mass. For examples : some of the 

 chemical classifications take only certain components into 

 account ; Fouque and Michel-Levy classify igneous rocks by 

 the character of that variable portion of the magmas consolidat- 

 ing during the second period ; Rosenbusch uses the phenocrysts 

 only, in certain parts of his system. 



The fundamental requirement that systems should be logical 

 in construction, with consistent and consequent application of 

 principles adopted, has been so commonly disregarded that a 

 summary of instances in point seems unnecessary. Some of the 

 most widely used systems of today are notably illogical as to 

 criteria, as has been pointed out. 



One of the most serious defects of modern classifications of 

 igneous rocks is a matter of bad logic, and to this defect the 

 writer wishes to allude once more. It is commonly admitted 

 that the chemical composition of these rocks is their most funda- 

 mental characteristic, and many authors would apparently be 

 glad to apply this character in classification. It is generally 

 stated, however, that the chemical is represented by the mineral 

 composition, and as the minerals are so prominent it is con- 

 venient to use them in system. But with no further discussion 

 it has been the universal plan to use the minerals in so limited a 

 qualitative way that they do not in fact express chemical com- 

 position except in a most crude and inadequate manner. This 

 procedure is purely and simply illogical, if the intention be to 

 represent chemical composition by the minerals of the rock. 



